• Korean Journal of Remote Sensing
• /
• v.37 no.6_1
• /
• pp.1573-1587
• /
• 2021

A Drone image is an ultra-high-resolution image that is several or tens of times higher in spatial resolution than a satellite or aerial image. Therefore, drone image-based remote sensing is different from traditional remote sensing in terms of the level of object to be extracted from the image and the amount of data to be processed. In addition, the optimal scale and size of data used for model training is different depending on the characteristics of the applied deep learning model. However, moststudies do not consider the size of the object to be found in the image, the spatial resolution of the image that reflects the scale, and in many cases, the data specification used in the model is applied as it is before. In this study, the effect ofspatial resolution and image size of drone image on the accuracy and training time of the semantic segmentation deep learning model of six wintering vegetables was quantitatively analyzed through experiments. As a result of the experiment, it was found that the average accuracy of dividing six wintering vegetablesincreases asthe spatial resolution increases, but the increase rate and convergence section are different for each crop, and there is a big difference in accuracy and time depending on the size of the image at the same resolution. In particular, it wasfound that the optimal resolution and image size were different from each crop. The research results can be utilized as data for getting the efficiency of drone images acquisition and production of training data when developing a winter vegetable segmentation model using drone images.

• Journal of Environmental Science International
• /
• v.23 no.2
• /
• pp.239-248
• /
• 2014

This study investigated the effects of soil depths and soil organic fertilizer application on the growth characteristics of Spiraea bumalda 'Gold Mound' in a extensive green roof system. The treatments were 3 soil depths (10, 15 and 25 cm) and 5 soil types in mixture of artificial soil and organic fertilizer. We measured plant height, leaf width, leaf length, number of flowers, visual quality and survival rate from March to October in 2011. The growing medium of 10 cm soil depth showed the highest plant growth in $A_1$ (amended soil 100%), and the lowest plant growth in $O_1A_4$ (organic fertilizer 20% + amended soil 80%) treatment. In case of 15 cm soil depth, Spiraea bumalda 'Gold Mound' showed a high leaf length and visual quality in $O_1A_2$(organic fertilizer 33% + amended soil 67%) treatment and high leaf width and number of flowers in $O_1$ (organic fertilizer 100%) treatment. $A_1$ treatment without organic fertilizer showed the lowest leaf length and poorest visual quality, and $O_1A_4$ treatment showed the lowest plant height and lowest number of flowers. At soil depth 25 cm, $O_1A_1$ (organic fertilizer 50% + amended soil 50%) treatment showed greater plant height, visual quality and number of flowers than other treatments. The leaf length and leaf width were more effective in $O_1$ treatment. $A_1$ treatment showed a relatively low leaf length, leaf width and visual quality. The higher the organic conditioner, the better the plant growth. And, survival rates of Spiraea bumalda 'Gold Mound' showed 92%, 88% and 76% at soil depths of 25 cm, 15 cm and 10 cm, respectively, in this a extensive green roof system. Therefore, the results showed that the growth of Spiraea bumalda 'Gold Mound' was affected by both soil quality and soil depth. Different optimal mixtures of organic fertilizer and amended soil were determined, depending upon soil depth.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.13 no.2
• /
• pp.95-109
• /
• 1993

In this study, the optimal configuration of arch structure has been tested by a decomposition technique. The object of this study is to provide the method of optimizing the shapes of both two hinged and fixed arches. The problem of optimal configuration of arch structures includes the interaction formulas, the working stress, and the buckling stress constraints on the assumption that arch ribs can be approximated by a finite number of straight members. On the first level, buckling loads are calculated from the relation of the stiffness matrix and the geometric stiffness matrix by using Rayleigh-Ritz method, and the number of the structural analyses can be decreased by approximating member forces through sensitivity analysis using the design space approach. The objective function is formulated as the total weight of the structures, and the constraints are derived by including the working stress, the buckling stress, and the side limit. On the second level, the nodal point coordinates of the arch structures are used as design variables and the objective function has been taken as the weight function. By treating the nodal point coordinates as design variable, the problem of optimization can be reduced to unconstrained optimal design problem which is easy to solve. Numerical comparisons with results which are obtained from numerical tests for several arch structures with various shapes and constraints show that convergence rate is very fast regardless of constraint types and configuration of arch structures. And the optimal configuration or the arch structures obtained in this study is almost the identical one from other results. The total weight could be decreased by 17.7%-91.7% when an optimal configuration is accomplished.

• Journal of Digital Convergence
• /
• v.11 no.5
• /
• pp.237-243
• /
• 2013

This study aims to examine the modified $A^2/O$ process is useful to reduce the environmental pollution caused by nutrient in wastewater. Specific results are as follows: The removal rate was evaluated at each time period, ie., 18h, 8h, 6h, and 3h after the reaction started. The anoxic rate was more than 94-97% from 18h to 6h but was less than 50% before 6h. Thus, the test of nitrification was done using 6h as the optimal anoxic retention time and the aerobic retention time set at 24h. When the flow change was 1:1, the average ammonia concentration inputted was $30mg/{\ell}$. Returned top nitric acid solution and the concentration of ammonia solution falling into the anoxic reactor was about 50％ of the initial concentration, and the flow change was 1:2, the concentration of ammonia falling into the anoxic reactor was about 62% of that of influxed ammonia. And the results of this study showed that the nitrogen removal rate can be improved by inputting untreated nitric acid and changing the flow of top nitrate solution using the modified $A^2/O$ method.

• Animal Bioscience
• /
• v.34 no.8
• /
• pp.1392-1402
• /
• 2021

Objective: The growth rate of pigs is related to differentiation and proliferation of muscle cells, which are regulated by growth factors and expression of growth-related genes. Thus, the objective of this study was to establish optimal culture conditions for Jeju black pig (JBP) muscle cells and determine the relationship of various factors involved in muscle growth with the proliferation of JBP muscle cells. Methods: Muscles were taken from the femur skeletal muscle of JBP embryos. After isolation of the muscle cells, cells were cultured in a 6-well plate under four different culture conditions to optimize culture conditions for JBP muscle cells. To analyze proliferation rate of JBP muscle cells, these muscle cells were seeded into 6-well plates at a density of 1.5×10⁵ cells per well and cultured for 3 days. Western blot and quantitative real-time polymerase chain reaction were applied to verify the myogenic differentiation 1 (MyoD) expression and growth-related gene expression in JBP muscle cells, respectively. Results: We established a muscle cell line from JBP embryos and optimized its culture conditions. These muscle cells were positive for MyoD, but not for paired box 7. The proliferation rate of these muscle cells was significantly higher in a culture medium containing bFGF and epidermal growth factor + basic fibroblast growth factor (EGF+bFGF) than that without a growth factor or containing EGF alone. Treatment with EGF and bFGF significantly induced the expression of MyoD protein, an important transcription factor in muscle cells. Moreover, we checked the changes of expression of growth-related genes in JBP muscle cells by presence or absence of growth factors. Expression level of collagen type XXI alpha 1 gene was changed only when EGF and bFGF were added together to culture media for JBP muscle cells. Conclusion: Concurrent use of EGF and bFGF increased the expression of MyoD protein, thus regulating the proliferation of JBP muscle cells and the expression of growth-related genes.

• Journal of Information Technology Applications and Management
• /
• v.19 no.1
• /
• pp.1-12
• /
• 2012

Data mining is the process of searching and analyzing large quantities of data for finding out meaningful patterns and rules. Artificial Neural Network (ANN) is one of the tools of data mining which is becoming very popular in forecasting the future values. Some of the areas where it is used are banking, medicine, retailing and fraud detection. In finance, artificial neural network is used in various disciplines including stock market forecasting. In the stock market time series, due to high volatility, it is very important to choose a model which reads volatility and forecasts the future values considering volatility as one of the major attributes for forecasting. In this paper, an attempt is made to develop two models - one using feed forward back propagation Artificial Neural Network and the other using Autoregressive Conditional Heteroskedasticity (ARCH) technique for forecasting stock market returns. Various parameters which are considered for the design of optimal ANN model development are input and output data normalization, transfer function and neuron/s at input, hidden and output layers, number of hidden layers, values with respect to momentum, learning rate and error tolerance. Simulations have been done using prices of daily close of Sensex. Stock market returns are chosen as input data and output is the forecasted return. Simulations of the Model have been done using MATLAB$^{(R)}$ 6.1.0.450 and EViews 4.1. Convergence and performance of models have been evaluated on the basis of the simulation results. Performance evaluation is done on the basis of the errors calculated between the actual and predicted values.

• The KIPS Transactions:PartB
• /
• v.16B no.4
• /
• pp.319-326
• /
• 2009

PSO(Particle Swarm Optimization) is an optimization algorithm in which simple particles search an optimal solution using shared information acquired through their own experiences. PSO applications are so numerous and diverse. Lots of researches have been made mainly on the parameter settings, topology, particle's movement in order to achieve fast convergence to proper regions of search space for optimization. In standard PSO, since each particle uses only information of its and best neighbor, swarm does not explore diverse regions and intended to premature to local optima. In this paper, we propose a new particle's movement strategy in order to explore diverse regions of search space. The strategy is that each particle moves according to relative weights of several better neighbors. The strategy of exploring diverse regions is effective and produces less local optimizations and accelerating of the optimization speed and higher success rates than standard PSO. Also, in order to raise success rates, we propose a strategy for checking whether swarm falls into local optimum. The new PSO algorithm with these two strategies shows the improvement in the search speed and success rate in the test of benchmark functions.

• Journal of Korea Multimedia Society
• /
• v.4 no.3
• /
• pp.230-237
• /
• 2001

Adaptive filters are widely used for acoustic echo canceler, adaptive equalizer and adaptive noise canceler. Coefficients of adaptive filters are updated by NLMS algorithm. However, Coefficients are misaligned by ambient noises when they are adapted by NLMS algorithm. In this Paper, a method determined the adaptation constant by low-pass filtered instantaneous gradient vector of LMS algorithm using orthognality principles of optimal filter is proposed. At initial states, instantaneous gradient vector, that is the cross-correlation of input signals and estimation error signals, has large value because input signals are remained in estimation error signals. When an adaptive filter is conversed, the cross-correlation will be close to zero. It isn's affected by ambient noises because ambient noises are uncorrelated with input signals. Determining adaptation constant with the cross-correlation, adaptive filters can be robust to ambient noises and the convergence rate doesn't slower As results of computer simulations, it is shown that the performance of proposed algorithm is betted than that of conventional algorithms.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.4
• /
• pp.921-929
• /
• 2011

This paper describes an analysis of the decoding performance and decoding convergence speed of layered LDPC(low-density parity-check) decoder for mobile WiMAX system, and the optimal design conditions for hardware implementation are searched. A fixed-point model of LDPC decoder, which is based on the min-sum algorithm and layered decoding scheme, is implemented and simulated using Matlab model. Through fixed-point simulations for the block lengths of 576, 1440, 2304 bits and the code rates of 1/2, 2/3A, 2/3B, 3/4A, 3/4B, 5/6 specified in the IEEE 802.16e standard, the effect of internal bit-width, block length and code rate on the decoding performance are analyzed. Simulation results show that fixed-point bit-width larger than 8 bits with integer part of 5 bits should be used for acceptable decoding performance.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.9 no.1
• /
• pp.33-40
• /
• 1989

In the design of reinforced concrete framed structures, which consist of various design variables, the objective and the constraint functions are formulated in complicated forms. Usually iterative methods have been used to optimize the design variables. In this paper, multilevel formulation is adopted, and design variables are selected in reduced numbers at each level, to reduce the iterative cycle and to accelerate the convergence rate. At level 1, elastic analysis is performed to get the upper and lower bounds of the redistributed design moments due to inelastic behavior of the frame. Then the design moments are taken as design variables and optimized at level 2, and the sizing variables are optimized at level 3. The optimization of redistributed moments is performed using the design sensitivity obtained at the level 2, and force approximation technique is used to reflect the variation of design variables in the lower level to the upper level. The design variables are selected in reduced numbers at each level, and the optimization formulation is simplified effectively. A cost function is taken as the objective function, and the constraints of the stress of the structures are derived from BSI CP 110 following limit state theory. Numerical examples are included to prove the effectiveness of the developed algorithm.